The best estimates put the traditional interview as still the most frequently used selection ´technique;´ used in more or less 60% of the time. <br> <p>Perhaps the Biggest Disconnect</p>
<p>Here is the ´disconnect.´ The best estimates put the traditional interview as still the most frequently used selection ´technique;´ used in more or less 60% of the time. And yet, when it comes to the validity of the traditional interview, just about every study in the last 30 years has found it at or near the bottom of the list.</p>
<p>There are other associated ´disconnects.´ For instance, despite all the talk about ´quality-of-hire,´ we still find that with most organizations only a minority of them collect and report ´quality-of-hire´ metrics-time-to-fill and cost-per-hire being much more popular metrics.</p>
<p>Interestingly, I have not seen an ROI model that pulls together all these factors. Here is what I am thinking about. Consider, say, that you have $1,000 to spend on improving your recruitment system. Would the ROI be best if that $1,000 were spent: (a) reducing time-to-fill, (b) reducing cost-per-hire, or (c) increasing the validity of selection tools?</p>
<p>I will admit that I do not have any data that directly address this question, but it seems that the greatest ROI would be in the use of more valid selection tools. This is simply based on the idea that the lifetime return that better employees will bring to the organization will usually dwarf selection costs.</p>
<p>So why are we reluctant to invest in increasing the validity of selection processes?</p>
<p>One reason might be that costs are easier to document and are all in by the time of hire; the value of a better hire, however, can only be established over a longer timeframe. So, if you want to look good, demonstrating reduced cost-per-hire is more immediate and more certain than demonstrating better quality-of-hire.</p>
<p>Another possibility may be that many don´t believe that the validity of selection processes could be improved that significantly. Despite what the data show, most hiring managers feel that they are able to make good selection decisions when they interview candidates. Therefore, there cannot be that much to be gained from more sophisticated selection processes.</p>
<p>So it becomes a matter of whether we trust the data or not. The data clearly suggest that the traditional interview is a poor predictor of future employee performance. But if you don´t trust the data, then the data, no matter how much there is, is just not persuasive.</p>
<p>Not trusting data is tricky. On the one hand, it is obvious that not all data should be trusted all the time. On the other hand, if we get in the habit of ´not trusting´ data we don´t agree with, the whole area of HR metrics and measurement can become dicey.</p>
<p>I am reminded of the book <i>Moneyball</i> by Michael Lewis. The whole point was that the real competitive advantage was when there was a discrepancy between ´accepted wisdom´ and ´what the data actually showed.´</p>
<p>The accepted wisdom appears to be that you create more value by reducing time-to-fill and/or cost-per-hire; the competitive advantage may well be on focusing on the validity of selection processes.</p>
